Methods for testing an image based occupant classification system

Abstract

A method for testing an image based occupant classification system includes identifying a plurality of disturbances for an image based occupant classification system and identifying a plurality of test occupants for a vehicle. The method further includes randomly selecting at least one disturbance from the plurality of disturbances and randomly selecting a test occupant from the plurality of test occupants.

1. A method for testing an image based occupant classification system comprising an image sensor, the method comprising:
 constructing a fault tree that includes:
 a plurality of first physical events that cause the image based occupant classification system to malfunction when the image sensor obtains a valid image; and 
 a plurality of second physical events that cause the image sensor to obtain an invalid image; 
 
 identifying a plurality of test occupants for a vehicle; 
 randomly selecting at least one of the first physical events from the fault tree; 
 randomly selecting at least one of the second physical events from the fault tree; 
 randomly selecting a test occupant from the plurality of test occupants; 
 physically introducing said randomly selected first physical event, said randomly selected second physical event, and said randomly selected test occupant to the image sensor at different times; 
 operating the image sensor to facilitate determination of an occupant characteristic by the image based occupant classification system for each of said randomly selected first physical event, said randomly selected second physical event, and said randomly selected test occupant; 
 determining whether the image based occupant classification system properly identified a characteristic of the test occupant for each of said randomly selected first physical event, said randomly selected second physical event, and said randomly selected test occupant; 
 calculating a failure rate of the image based occupant classification system; and 
 reconfiguring an algorithm of the image based occupant classification system when the failure rate is above a predetermined level. 
 
     
     
       2. The method of  claim 1  further comprising identifying a driving condition and introducing the driving condition to the image based occupant classification system. 
     
     
       3. The method of  claim 2  wherein introducing the driving condition comprises operating a vehicle according to the driving condition, wherein the vehicle includes the image based occupant classification system. 
     
     
       4. The method of  claim 1  wherein constructing the fault tree further comprises compiling a list of the plurality of first physical events. 
     
     
       5. The method of  claim 1  wherein constructing the fault tree further comprises compiling a list of the plurality of second physical events. 
     
     
       6. The method of  claim 1  wherein identifying the plurality of test occupants comprises compiling a list of the plurality of test occupants. 
     
     
       7. The method of  claim 1  wherein randomly selecting at least one of the first physical events randomly selecting at least one of the second physical events, and randomly selecting a test occupant comprises employment of a randomizer. 
     
     
       8. The method of  claim 1  wherein the plurality of first physical events comprises:
 a plurality of lighting issues that cause the image based occupant classification system to malfunction when the image sensor obtains a valid image; 
 a plurality of blocked view events that cause the image based occupant classification system to malfunction when the image sensor obtains a valid image; and 
 a plurality of occupant out of view events that cause the image based occupant classification system to malfunction when the image sensor obtains a valid image. 
 
     
     
       9. The method of  claim 1  wherein the plurality of second physical events comprises:
 a plurality of poor field of view issues that cause the image sensor to obtain an invalid image; and 
 a plurality of lens faults that cause the image sensor to obtain an invalid image. 
 
     
     
       10. A method for testing an image based occupant classification system comprising an image sensor, the method comprising:
 constructing a fault tree that includes:
 a plurality of lighting issues that cause the image based occupant classification system to malfunction when the image sensor obtains a valid image; 
 a plurality of blocked view events that cause the image based occupant classification system to malfunction when the image sensor obtains a valid image; 
 a plurality of occupant out of view events that cause the image based occupant classification system to malfunction when the image sensor obtains a valid image; 
 a plurality of poor field of view issues that cause the image sensor to obtain an invalid image; and 
 a plurality of lens faults that cause the image sensor to obtain an invalid image; 
 
 identifying a plurality of test occupants for a vehicle, the test occupants comprising an empty seat, a child seat, and a seated adult; 
 randomly selecting one of the plurality of lighting issues, one of the plurality of blocked view events, one of the plurality of occupant out of view events, one of the plurality of poor field of view issues, and one of the plurality of lens faults from the fault tree; 
 randomly selecting a test occupant from the plurality of test occupants; 
 physically introducing the randomly selected lighting issue, the randomly selected blocked view event, the randomly selected occupant out of view event, the randomly selected poor field of view issue, the randomly selected lens fault, and the randomly selected test occupant to the image sensor at different times; 
 operating the image sensor to facilitate determination of an occupant characteristic by the image based occupant classification system for each of the randomly selected lighting issue, the randomly selected blocked view event, the randomly selected occupant out of view event, the randomly selected poor field of view issue, the randomly selected lens fault, and the randomly selected test occupant; 
 determining whether the image based occupant classification system properly identified a characteristic of the test occupant for each of the randomly selected lighting issue, the randomly selected blocked view event, the randomly selected occupant out of view event, the randomly selected poor field of view issue, the randomly selected lens fault, and the randomly selected test occupant; 
 calculating a failure rate of the image based occupant classification system; and 
 reconfiguring an algorithm of the image based occupant classification system when the failure rate is above a predetermined level. 
 
     
     
       11. The method of  claim 10  further comprising identifying a driving condition and introducing the driving condition to the image based occupant classification system. 
     
     
       12. The method of  claim 11  wherein introducing the driving condition comprises operating a vehicle according to the driving condition, wherein the vehicle includes the image based occupant classification system. 
     
     
       13. The method of  claim 10  wherein identifying the plurality of test occupants comprises compiling a list of the plurality of test occupants. 
     
     
       14. The method of  claim 10  wherein randomly selecting one of the plurality of lighting issues, one of the plurality of blocked view events, one of the plurality of occupant out of view events, one of the plurality of poor field of view issues, and one of the plurality of lens faults from the fault tree, and randomly selecting a test occupant comprises employment of a randomizer.